One dangling link anywhere in a repository, often in the author's own
agent settings and unrelated to any plugin, made every plugin in it
uninstallable. Such links (broken, absolute, escaping the source, into
.git, or closing a cycle) are now left out of the digest and snapshot
and reported as a discovery warning. They are still never followed, and
sources without them keep the same digest, so existing bindings are
unaffected. Special files and the size limits still fail the source.
Plugin installs could only be checked against real Agents on the host.
The CLIs are installed on the dev-home volume so image rebuilds keep
them, and relinked into /usr/local/bin on every start so non-login
docker exec finds them. A failed install warns without blocking start.
One broken or external entry used to fail the whole marketplace, and
valid Claude plugins were rejected because skillshare expected more than
Claude does. Checked against the real Claude and Codex CLIs:
- A bad entry now marks only its own row, with a translatable reason.
- An external entry yields to a catalog that points inside the source.
- Claude plugins without plugin.json are read from their catalog entry
and default folders, including a root SKILL.md as the only skill.
- Claude entries keep their own definition (strict, skills, lspServers,
version, ...) in the catalog written for Claude, so entries sharing a
folder no longer install identical content.
- When catalog and manifest names differ, Claude still installs under
the catalog name; Codex refuses, so other Agents are blocked.
- When catalogs place the same plugin in different folders per Agent,
each Agent installs from its own folder.
- The add dialog no longer preselects a plugin that cannot be installed.
Pi prints its version to stderr, so the host showed an empty version.
The fallback applies only to --version with empty stdout, so commands
whose stdout is parsed never see stderr noise.
Local skills in a target showed their internal synced/<org>_<uuid> path,
and they inflated the always-loaded and on-demand totals with skills
skillshare does not manage. Local rows now show their folder name (full
path on hover), and a toggle, remembered per browser, limits the list
and the totals to managed skills.
The rail layout gives its list a viewport-height scroll box, which suits the
MCP page but nests a second scroll area under the project page's header and
tabs, and its narrow padding clipped Playful borders.
Skills and agents settings are pointers shared with the live config, so a
request rejected on a later field had already applied earlier ones, and the
next successful save wrote them to disk.
Tools disagree on agent frontmatter and some need another format entirely,
so an agent written for Claude is misread elsewhere. An `extension` on a
target's agents block reuses the extras transform mechanism to convert each
agent during sync.
Converted targets hold transformed output, so collect never pulls them back
into the source. Leftover symlinks from merge or symlink mode are removed
rather than written through, and a failed conversion drops the old output
so an unconverted agent never stays active.
The bundled opencode-agents extension keeps only fields OpenCode documents
and refuses agents that restrict their tools, since dropping those fields
would widen the agent's access.
Refs #267
Refs #241
The pages still said a disabled entry cannot target Claude Code, and that the
field is project mode only. Both are lifted: the entry is valid under an
mcp.projects root as well, and for Claude Code the switch goes to
~/.claude.json, where Claude Code keeps its per-project off list, leaving the
server definitions in that file alone. Only the global mcp.servers, where no
project is in scope, still refuses it.
The dashboard sentence was stale from an earlier change: the switch is its own
"Turn off a global server" button beside "Add server", shown in project mode
and on a project's MCP tab, not an option inside the add form.
All four locales follow, naming the buttons as the dashboard shows them there.
Refs #286
tsc -b fails on three counts, which the dashboard image build turns into a red
CI. Two came in with the mcp.projects off switch, one is older.
MCPRemoveDialog's inScope predicate now takes a change rather than a path, but
the type it names was never imported, and MCPPage still handed it the old
path-based predicate. That call site has to ask the change for its scope
anyway: a project's Claude Code off switch is written to ~/.claude.json, a
global path, so matching on the path alone would count it as a global change.
ConfigPage passed its tab straight to AssistantPanel, whose mode has never
accepted the extensions tab. Only the branch that excludes that tab renders
the panel, so the value cannot occur; say so at the call site.
Claude Code keeps its per-project off list in ~/.claude.json, the same file
its user-scope servers live in. The plan keyed desired entries by file path
alone, so one file could carry only one native target, and mcp.projects had
to refuse a disabled entry for Claude Code rather than write that list.
Key the plan by file and target instead. Apply already rereads the file per
plan, and sectionDigest hashes only the target's own section, so two plans
on ~/.claude.json fold in order without further machinery.
A change now carries the mcp.projects root it belongs to. Its path no longer
says, because a project's off switch is written outside that project folder,
and the dashboard needs the root to keep the switch out of the global server
list and to show a pending badge on the project it came from.
Codex stays refused. It merges the project file field by field, so a switch
for a server its global config lacks leaves an entry with no transport and
stops Codex loading its config at all.
Alongside this, the dashboard gets a named "Turn off a global server" action,
so adding a switch and adding a server are two choices up front rather than
one form that asks afterwards. Two older glitches go with it: a project's
target toggles sprang back while the list was still empty, and the expanded
config editor stayed open when a tab arrived from the URL.
Refs #286
The inline editor shares the row with the file list and assistant panel,
which leaves little room for long YAML. An Expand button opens the same
editor and panel in a near-fullscreen dialog. The dialog ignores backdrop
clicks and Escape so an edit in progress is never dismissed by accident.
Teams need to know that skills.lock.json is meant to be committed, which
commands move a pin, and how it relates to the config.yaml manifest.
Covers project-skills, install, update, uninstall, declarative-manifest
and project-setup, in English and the four translated locales.
config.yaml records which skills a project wants, but not which commit
each one resolved to, so two teammates running `install -p` a week apart
could end up with different skill contents.
Project mode now writes .skillshare/skills.lock.json next to the config.
`install -p` checks out the locked commit for skills and tracked repos,
and brings an already installed skill to the lock when they differ.
`update -p` and the dashboard move a pin only for skills whose commit
actually changed, so a stale local copy cannot roll the team back, and
uninstall drops the pin. A locked value is only honoured when it is a
full SHA for the same source, since it ends up as a git argument.
A skill installed into a group had its metadata written to the group
folder instead of the skills root, so it never reached config.yaml and
teammates running `install -p` did not get it.
Two paths left the skills root unset and fell back to the parent of the
skill: `install <git source> --into <group>`, because only Install
derived the root and InstallFromDiscovery did not, and config installs
of an entry with `group:`, in both project and global mode. Both now
resolve the root the same way.
The analyze numbers were drifting from what tools really see. Symlink-mode
targets expose the whole source folder, so .skillignore'd skills still load
but were left out; merge/copy targets often hold skills the user dropped in
by hand that skillshare never counted; and chars/4 undercounted CJK text by
roughly four times.
Route the CLI, TUI, web handler and sync summary through one
sync.TargetSkills so the three copies of the filter logic cannot disagree
again. Disabled skills stay in the discovery with Disabled=true and are
priced for symlink targets; unmanaged folders in a target are reported as
Local. Token estimates now charge one token per wide rune.
The dashboard keeps the chosen target in the URL so target and project
pages can deep-link into the analysis, lists project targets after own
targets with their tool icon, and tags local/disabled rows. Select gives
single-select icons a key to silence a React warning.
Installing from a known repo URL is the common path, so it now leads the
tab list and is where the dialog lands by default. The ?install=search
link still opens the search tab.
Keeps the sync reference in every locale in step with the English page,
which now explains why project sync removes leftovers after a target's
default path moves.
Refs #135
Both tools now document ~/.agents/skills and <project>/.agents/skills as
their skills locations and keep the old directories only for backward
compatibility. Writing to the old path next to `universal` makes them read
every skill from two places.
Global configs pin absolute paths, so they are unaffected and keep their
overlap warning. Project configs store target names only, so an existing
project follows the new default on its next sync while the links it created
in .goose/skills or .openhands/skills would stay behind and show up as
duplicates. Project sync now removes those leftovers from also_scans
directories that no configured target writes to.
That directory belongs to another tool, so the cleanup runs the prune with
ManagedOnly: only symlinks into the source and manifest-tracked folders are
removed. The flat-name heuristic would otherwise delete hand-made folders
such as `_drafts`. Targets with an explicit path are skipped, and --dry-run
only previews.
Refs #135
The Antigravity CLI (agy) reads ~/.gemini/antigravity-cli/skills and
never the app's ~/.gemini/config/skills, so treating it as an alias of
antigravity synced skills to a directory it does not read. Give it its
own targets.yaml entry with the correct path.
A path-less `antigravity-cli:` config entry also failed validation,
because DefaultTargets() is not alias-aware; as a real target it now
resolves to the default path.
- Fallback fetch pulled only the default branch because a depth-1 clone
is single-branch; fetch all heads and tags so an abbreviated SHA off
the default branch resolves.
- Move the --track guard from CLI flag parsing into InstallTrackedRepo
so the Web UI is covered too, and reject tags as well as SHAs: both
leave HEAD detached with nothing for `git pull` to follow.
- An all-hex branch or tag name (e.g. 20240915) was treated as a SHA in
check and never queried; look the ref up first and fall back to the
literal SHA only when the remote has no such ref. Full 40-char SHAs
still skip the network.
- Rename the git package's 7-char shortHash to abbrevHash to avoid
clashing with install's 12-char helper.
Refs #281
Codex documents ~/.agents/skills as its user-level skills path and marks
~/.codex/skills as deprecated, while still reading both. It does not merge
skills that share a name, so the default init result (codex + universal)
showed every skill twice in Codex.
The codex target now defaults to the shared directory, keeps the old path
in also_scans, and is detected through ~/.codex. A fresh init therefore sets
up only `universal` for Codex and writes each skill once. Existing configs
keep their pinned path and still get the doctor overlap warning that
suggests removing the codex target.
Refs #135
Teams with supply-chain review need reproducible installs, but --branch
only worked for branch names because git clone --branch rejects SHAs.
Clone the default branch and detach at the pinned commit (targeted
depth-1 fetch, falling back to --unshallow for abbreviated SHAs or
servers that refuse arbitrary wants). Resolve tags in check via
refs/tags and the peeled ^{} commit, and short-circuit SHA pins without
a network round-trip since they can never move. Reject --track with a
SHA: a tracked repo must follow a branch to pull.
Closes#281
Several tools now document the shared ~/.agents/skills directory as their
primary skills path. Moving a built-in target there would break three things
that assume a tool-specific path: init and the dashboard detect an installed
tool through the parent of its skills path, and the overlap checks only
consult also_scans, so configs pinned to the old path would silently lose
their duplicate-skill warning.
Add an optional `detect` install dir per target and derive overlap checks
from every path a runtime reads (also_scans plus the built-in default),
skipping the target's own configured path. No default changes here, so
behaviour is unchanged until a target opts in.
Refs #135
Global skills already reach every project, so the docs lead with when
`projects` is not needed and compare it with plain global targets and
project mode. The recipe, config reference, MCP and UI command pages
are updated in English and the four translated locales.
Refs #286
People with many repositories had to choose between listing every
project skills folder as a custom target or running project mode in
each repo. A `projects:` key in the global config now declares a folder
once with the tools used there, and sync treats each one as a
`<project>@<tool>` target with its own mode and filters.
- Expand `projects:` into virtual targets at load time and strip them
on save; tools that share a project skills folder are written once
- Skip a project whose folder is missing with a warning instead of
failing the whole sync; `collect` and `target` ignore project targets
- Add /api/projects (list, save, remove, convert) and detect ordinary
targets that already point inside a project so they can be converted
- Add a Projects page to the dashboard with Skills, Agents and MCP tabs
per project, replacing the Projects tab on the MCP page
- Extend Select with a multi-select mode for picking a project's targets
Refs #286
Saving config.yaml already ran it through formatYaml, so the editor
could reformat the file without the user ever seeing the result. Expose
that same normalizer as a Beautify button on the config.yaml tab: it
writes the formatted YAML back into the buffer, so the change is visible
and revertable before saving.
also_scans drives the doctor overlap warnings, but it only covered nine
targets and had drifted from upstream. Re-checked every entry against
vendor docs or vendor source and applied one rule: list a path only when
the tool scans it by default. Opt-in and experimental scans stay out.
- kimi: drop ~/.agents/skills. Kimi reads it only when
~/.config/agents/skills is absent, and skillshare always creates the
latter, so the warning was a false positive.
- amp, kimi, cursor, codex: add documented paths that were missing.
- opencode, goose, copilot, crush, droid, pi, cline, commandcode,
deepagents, kilocode, openclaw, openhands, kode: add also_scans for the
first time. Users pairing these with claude or universal got duplicate
listings with no doctor warning.
Add a guard test so a target never lists its own primary path.
Refs #135
The cross_target_discovery suggestion pointed users at removing the
writer target (e.g. universal). That target owns a shared directory, so
removing it also hides skills from every other tool that reads the same
path. Removing the scanning target (e.g. codex) only affects that one
runtime, which still receives the content through the shared directory.
Reword the suggestion and the matching docs resolution in all locales.
Refs #135
The mcp reference gains sections for managing several projects from the
global config, the dashboard's Projects tab and Defaults, and Pi direct
tools. It states that a project lists only what differs: a global server
already loads in every project because the Agent merges its global and
project files, a point readers arriving at that section kept missing.
A new recipe, many projects from one config, shows skills and MCP servers
together, since custom path targets already solve the skills half.
The built-in skill's MCP reference is rewritten around the same material so
an Agent editing config.yaml for a user knows these rules.
All of it is mirrored in zh-Hant, zh-Hans, ja and ko.
Refs #286
mcp.projects had no editor, so using it meant hand-writing YAML, which is
the friction the feature set out to remove.
In global mode the MCP page gains a Projects tab. Adding a project takes a
folder and its targets. A project's page lists every global server with a
switch that saves or removes a disabled entry for the Agents that support a
per-project switch, and below it the servers that exist in that project
only. Claude Code is left out of those switches because mcp.projects cannot
turn a server off for it.
A Defaults section on the Servers tab edits mcp.targets and mcp.directTools,
which had no write path before. Direct tools is shown only once Pi is among
the effective targets, since only pi-mcp-adapter reads it; the same rule
applies in the add-project dialog and on the project page.
Every settings save sends targets and directTools together, because the API
replaces both and a value left out would be cleared.
Refs #286
Project mode keeps MCP settings in each project's .skillshare/config.yaml and
is synced from inside that folder. Someone who works across many projects had
to repeat that per folder and could not see or sync them in one place.
mcp.projects in the global config lists project roots, each with its own
targets, servers and directTools. One sync plans the global files and every
root together, and removing a root cleans up what was written there. It is
refused in a project config, and disabled cannot target Claude Code there
because its off list shares ~/.claude.json with the global servers.
directTools is now a first-class setting for pi-mcp-adapter: per server
(--direct-tools on mcp add/edit), as a default under mcp, or per project.
Before this it could only be hand-edited into Pi's file.
Mutations accept a project root and a settings block so the dashboard can
edit all of this. Saving re-encodes only the nodes that changed, so other
projects keep their YAML as written, anchors, aliases and a leading ~ in the
key included. expandHome also accepts ~\ so Windows paths resolve.
Refs #286
Completes zh-Hant, zh-Hans, ja, and ko translation of the final 50
pages (how-to/, learn/, reference/targets/, reference/appendix/,
reference/filtering.md, reference/index.md, troubleshooting/),
bringing all four locales to full parity with the English docs tree.
Anchor ids added in the previous commit are propagated to every
translated heading they land on, and one zh-Hans same-file link that
pointed at the pre-translation auto-slug is repointed at the pinned
id.
- how-to/advanced/security.md: pin {#periodic-scanning} on 'Periodic
Scanning', linked from this same page by anchor fragment.
- how-to/advanced/docker-sandbox.md: pin {#dev-profile} on 'Dev
Profile', same reason.
- how-to/daily-tasks/organizing-skills.md: pin
{#install-directly-into-folders} on that heading, same reason.
Without an explicit id, translating the heading text changes its
auto-generated slug and breaks the in-page link once translated.
- how-to/daily-tasks/creating-skills.md: the nested code-fence example
under 'Include examples' used the same 3-backtick fence for both the
outer block and the inner ```python block, so the inner fence's
info string didn't count as a closer but the following bare fence
did, silently truncating the outer block and leaving a dangling
opener for the rest of the file. Widened the outer fence to 4
backticks.
These headings mix prose with technical terms (e.g. '.skillignore
(Optional)', 'Git Authentication', 'Target include/exclude filters')
and are referenced by anchor fragment from other pages. Translating
the prose changes the auto-generated slug and breaks the link, so
pin the current slug explicitly before the next translation batch.
The unanchored patterns matched any nested file with that basename
(case-insensitively), so every translated understand/agents.md under
website/i18n/ and website/CLAUDE.md were silently excluded from every
commit since they were created. This broke the ja/ko/zh-Hans/zh-Hant
Pages build in CI: relative links to ./agents.md resolved to a file
that didn't exist in the checked-out tree, even though it existed
locally. Anchoring the patterns with a leading slash keeps them
scoped to the repo-root personal instruction files they were meant
for, and recovers the four missing translations plus website/CLAUDE.md.
Sets up Docusaurus i18n for Traditional Chinese, Simplified Chinese,
Japanese, and Korean: locale config, navbar/footer/sidebar UI chrome,
and translated content for intro, getting-started, understand, and
reference/commands. Untranslated pages fall back to English.
Pins explicit heading anchors ({#id}) on English source pages that are
linked to by fragment, so translated headings don't change the anchor
id and break cross-page links. Converts a handful of relative .md
links that crossed from an untranslated page into a now-translated
one, since Docusaurus resolves those per-locale and they'd otherwise
404 outside English.
After the previous fix the dashboard failed fast but showed the whole
CLI transcript ending in "sudo: a password is required", which does not
tell the user what to do next.
Probe `sudo -n true` before re-executing and return a message naming the
terminal command. The server maps that to an `upgrade.needs_sudo` error
code so the update dialog and the Doctor page can show a translated
sentence in every locale.
Also bound the release download with a client timeout; http.Get had
none, so a stalled connection could hang an upgrade indefinitely.
The dashboard's "Update now" spawns `skillshare upgrade --force` with no
stdin. When the binary lives in a root-owned directory the command
re-execs through sudo, which prompts on /dev/tty rather than stdin, so
the dialog spun until the 10 minute timeout waiting for a password
nobody could see.
Pass -n to sudo when stdin is not a terminal: cached credentials and
NOPASSWD setups still upgrade, everything else errors immediately and
the dialog can point the user at the terminal command.